Sanction teardown · CC New York, USA · 2026-01-15
Corst v. Mushailov
What happened
In CC New York, USA, a filing relied on an unnamed/unconfirmed AI tool to help draft legal argument. The court identified the following problems with the citations in that filing:
-
Fabricated (Case Law)Citation “Matter of Knapp v Knapp, 225 AD2d 1010 (4th Dept 1996)” could not be located under that name and instead led to a different matter (Orange Steel Erectors v Newburgh Steel Products Inc.); court found it defective/unverified.
-
Fabricated (Case Law)Citation “Matter of Negron v St. Charles Hosp. & Rehab. Ctr., 121 AD3d 1427 (2d Dept 2014)” led to a different matter (People v Rojas) and did not support counsel's argument; court could not find the cited matter.
-
Fabricated (Case Law)Citation “Matter of Polizzi v Polizzi, 226 AD2d 581 (2d Dept 1996)” did not correspond to that matter and led to a different case (Epes v Healey); court found it defective/unverified.
-
Misrepresented (Case Law)Citation “Wells Fargo Bank, N.A. v Estwick, 160 AD3d 911 (2d Dept 2018)” exists but the opinion does not support the proposition asserted by counsel; misrepresentation of precedent.
Which AI tool
an unnamed/unconfirmed AI tool. Note: Charlotin's public database records tool attribution only where a court order, brief, or reporting on the matter states it explicitly; "unidentified" or "implied" means the record indicates AI use but does not name a specific product — we do not guess.
Outcome
Apology from Counsel
How Citation Safe would have caught this
Citation Safe runs three deterministic layers before a brief is filed: (1) does the citation exist against CourtListener's database of published opinions, (2) if quoted, does that exact language appear in the source, (3) does the cited case actually support the proposition it is cited for. Fabricated case citations fail Layer 1. Fabricated or misattributed quotations fail Layer 2 even when the underlying case is real. Misrepresented holdings — a real case cited for a proposition it does not support — are the target of Layer 3. None of these checks involve asking another language model whether the citation looks right; they are lookups and text-matches against the actual source, which is why a hallucinated citation has to survive a direct lookup against the authoritative source — not another model's opinion — to earn a VERIFIED stamp; our measured false-verify rate is published live at /quality.
Check a brief before you file it → · See our live false-verify rate
Source: https://www.damiencharlotin.com/documents/2336/09_-_Corst_v_Mushailov.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).